Keep up with tech in 5 minutes
Get the free daily email with summaries of the most interesting stories in startups 🚀, tech 📱, and programming 💻!
Join 1,600,000 readers for one daily email
Repeating Prompts (1 minute read)
When not using reasoning, repeating the input prompt improves performance for popular models without increasing the number of generated tokens or latency. It is interesting that tricks like this are still possible despite the amount of work being put into improving large language models. The discovery proves how much room for improvement there still is for current models.
Feb 20 | Tech
White-Collar Apocalypse Isn't Around the Corner—But AI Has Already Fundamentally Changed the Economy (20 minute read)
The productivity gains from AI are real, but it won't be catastrophic, as it will just devalue mechanical skills while elevating judgment skills.
Feb 20 | Product
AI's not gonna kill SaaS, but it will force some changes (7 minute read)
AI will pressure SaaS, but it is not killing it. Large vendors still hold durable advantages in enterprise infrastructure and proprietary data, even as AI reshapes pricing and growth dynamics.
Feb 20 | Infosec
Clinejection — Compromising Cline's Production Releases just by Prompting an Issue Triager (13 minute read)
A prompt-injected GitHub issue title could drive Cline's Claude-based triage bot to execute arbitrary commands in CI, then use GitHub Actions cache poisoning to hijack nightly build workflows and steal VS Code Marketplace, OpenVSX, and npm publishing tokens. This chain could have enabled a massive supply-chain attack on millions of developers via malicious extension or CLI updates.
